Skripts as a clustered resource vs Procluster

Posted by Valeriy Bashkatov on 11-May-2011 10:41

Hello,

Tell me please, what are the advantages of PROCLUSTER (OpenEdge Clusters)? Instead of PROCLUSTER I can use scripts as a clustered resource to start the database after switching to the second cluster node. Then why need PROCLUSTER?

I want to understand the difference.

Regards,

Valeriy

OpenEdge RDBMS - Forum

All Replies

Posted by gus on 12-May-2011 12:33

Yes, you can certainly write your own scripts, and some people have done it, but it does take some effort. The goal of Procluster was to make it easy to use the failover capabilities and for its use to be largely operating system independent.

Posted by Valeriy Bashkatov on 16-May-2011 10:41

Thank you, Gus!
I have another question.

How to run an additional login-brokers (-m3) for the clustered database on HP-UX 11 (through procluster)?

Posted by Valeriy Bashkatov on 19-May-2011 06:08

>>How to run an additional login-brokers (-m3) for the clustered database on HP-UX 11 (through procluster)?

Hm.. Nobody know answer on this question?

P.S.

login-brokers = secondary broker

Posted by gus on 19-May-2011 15:41

It is possible to do by amending the scripts that are generated by procluster. I have to go research the details because I don't know the details involved.

Posted by Valeriy Bashkatov on 08-Jun-2011 05:26

Hello!

Somebody uses EVA Continuous Access Replication to replicate OpenEdge database with node A to node B?
Will the database to work after such a replication and after switching to node B?
The scheme of metroklaster attached to this post.

Posted by Valeriy Bashkatov on 08-Jun-2011 05:38

What problems may arise with us, if we replicate the database using "EVA Continuous Access Replication"?

Posted by Valeriy Bashkatov on 22-Jun-2011 05:33

We successfully tested the OpenEdge, which is replicated using HP CAR. Database good start after switching on the second node. The test was run with a heavy load on the database that generated using the ATM-test 100, 500, 1000, 2000 and 2500 users.

Failure of the primary node has emitted power down the server. The switching process takes no more than two or three minutes. After completion of Crash Recovery Database successfully started and continued to work.

Summary: OpenEdge RDBMS is running without errors in the HP MetroCluster + HP Continuous Access Replication.

But just in case we have to be parallel to support the work of OE Replication and After-Imaging database.

Thanks to everyone who helped me!

This thread is closed